| { | |
| "best_metric": 0.810126582278481, | |
| "best_model_checkpoint": "deit-base-distilled-patch16-224-55-fold2/checkpoint-101", | |
| "epoch": 85.71428571428571, | |
| "eval_steps": 500, | |
| "global_step": 300, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.8571428571428571, | |
| "eval_accuracy": 0.5443037974683544, | |
| "eval_loss": 0.7763431072235107, | |
| "eval_runtime": 1.1332, | |
| "eval_samples_per_second": 69.717, | |
| "eval_steps_per_second": 2.647, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.620253164556962, | |
| "eval_loss": 0.6779789328575134, | |
| "eval_runtime": 1.1183, | |
| "eval_samples_per_second": 70.64, | |
| "eval_steps_per_second": 2.683, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 2.252147674560547, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.721, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "eval_accuracy": 0.5316455696202531, | |
| "eval_loss": 0.6954271793365479, | |
| "eval_runtime": 1.1555, | |
| "eval_samples_per_second": 68.371, | |
| "eval_steps_per_second": 2.596, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.620253164556962, | |
| "eval_loss": 0.6370497941970825, | |
| "eval_runtime": 1.1498, | |
| "eval_samples_per_second": 68.706, | |
| "eval_steps_per_second": 2.609, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 4.857142857142857, | |
| "eval_accuracy": 0.5949367088607594, | |
| "eval_loss": 0.6104753017425537, | |
| "eval_runtime": 1.1377, | |
| "eval_samples_per_second": 69.439, | |
| "eval_steps_per_second": 2.637, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 5.714285714285714, | |
| "grad_norm": 3.993246555328369, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.6207, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.6835443037974683, | |
| "eval_loss": 0.5798102021217346, | |
| "eval_runtime": 1.1138, | |
| "eval_samples_per_second": 70.93, | |
| "eval_steps_per_second": 2.694, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 6.857142857142857, | |
| "eval_accuracy": 0.7468354430379747, | |
| "eval_loss": 0.5703896880149841, | |
| "eval_runtime": 1.1441, | |
| "eval_samples_per_second": 69.05, | |
| "eval_steps_per_second": 2.622, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7088607594936709, | |
| "eval_loss": 0.5878576636314392, | |
| "eval_runtime": 1.1443, | |
| "eval_samples_per_second": 69.035, | |
| "eval_steps_per_second": 2.622, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 8.571428571428571, | |
| "grad_norm": 8.652019500732422, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5427, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 8.857142857142858, | |
| "eval_accuracy": 0.6582278481012658, | |
| "eval_loss": 0.6726908087730408, | |
| "eval_runtime": 1.1277, | |
| "eval_samples_per_second": 70.055, | |
| "eval_steps_per_second": 2.66, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.6962025316455697, | |
| "eval_loss": 0.5840915441513062, | |
| "eval_runtime": 1.1335, | |
| "eval_samples_per_second": 69.694, | |
| "eval_steps_per_second": 2.647, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 10.857142857142858, | |
| "eval_accuracy": 0.6962025316455697, | |
| "eval_loss": 0.605922520160675, | |
| "eval_runtime": 1.1431, | |
| "eval_samples_per_second": 69.11, | |
| "eval_steps_per_second": 2.624, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 11.428571428571429, | |
| "grad_norm": 6.731996059417725, | |
| "learning_rate": 4.814814814814815e-05, | |
| "loss": 0.4775, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.6075949367088608, | |
| "eval_loss": 1.0271003246307373, | |
| "eval_runtime": 1.1385, | |
| "eval_samples_per_second": 69.392, | |
| "eval_steps_per_second": 2.635, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 12.857142857142858, | |
| "eval_accuracy": 0.7088607594936709, | |
| "eval_loss": 0.6412355303764343, | |
| "eval_runtime": 1.1419, | |
| "eval_samples_per_second": 69.181, | |
| "eval_steps_per_second": 2.627, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.6582278481012658, | |
| "eval_loss": 0.8064222931861877, | |
| "eval_runtime": 1.1524, | |
| "eval_samples_per_second": 68.552, | |
| "eval_steps_per_second": 2.603, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 14.285714285714286, | |
| "grad_norm": 9.111239433288574, | |
| "learning_rate": 4.62962962962963e-05, | |
| "loss": 0.4961, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 14.857142857142858, | |
| "eval_accuracy": 0.6582278481012658, | |
| "eval_loss": 0.5599769949913025, | |
| "eval_runtime": 1.1323, | |
| "eval_samples_per_second": 69.767, | |
| "eval_steps_per_second": 2.649, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.6708860759493671, | |
| "eval_loss": 0.5889371633529663, | |
| "eval_runtime": 1.1493, | |
| "eval_samples_per_second": 68.735, | |
| "eval_steps_per_second": 2.61, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 16.857142857142858, | |
| "eval_accuracy": 0.6835443037974683, | |
| "eval_loss": 0.8380601406097412, | |
| "eval_runtime": 1.1402, | |
| "eval_samples_per_second": 69.284, | |
| "eval_steps_per_second": 2.631, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 17.142857142857142, | |
| "grad_norm": 9.676458358764648, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 0.4391, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.6962025316455697, | |
| "eval_loss": 0.6724734902381897, | |
| "eval_runtime": 1.1265, | |
| "eval_samples_per_second": 70.13, | |
| "eval_steps_per_second": 2.663, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 18.857142857142858, | |
| "eval_accuracy": 0.7215189873417721, | |
| "eval_loss": 0.5349910855293274, | |
| "eval_runtime": 1.1568, | |
| "eval_samples_per_second": 68.293, | |
| "eval_steps_per_second": 2.593, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 5.236982822418213, | |
| "learning_rate": 4.259259259259259e-05, | |
| "loss": 0.413, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.7088607594936709, | |
| "eval_loss": 0.6033266186714172, | |
| "eval_runtime": 1.1208, | |
| "eval_samples_per_second": 70.483, | |
| "eval_steps_per_second": 2.677, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 20.857142857142858, | |
| "eval_accuracy": 0.6835443037974683, | |
| "eval_loss": 0.7280053496360779, | |
| "eval_runtime": 1.1275, | |
| "eval_samples_per_second": 70.064, | |
| "eval_steps_per_second": 2.661, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_accuracy": 0.7341772151898734, | |
| "eval_loss": 0.6081877946853638, | |
| "eval_runtime": 1.1526, | |
| "eval_samples_per_second": 68.541, | |
| "eval_steps_per_second": 2.603, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 22.857142857142858, | |
| "grad_norm": 3.7129807472229004, | |
| "learning_rate": 4.074074074074074e-05, | |
| "loss": 0.336, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 22.857142857142858, | |
| "eval_accuracy": 0.759493670886076, | |
| "eval_loss": 0.6529555916786194, | |
| "eval_runtime": 1.1446, | |
| "eval_samples_per_second": 69.021, | |
| "eval_steps_per_second": 2.621, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_accuracy": 0.7088607594936709, | |
| "eval_loss": 0.6921886205673218, | |
| "eval_runtime": 1.1486, | |
| "eval_samples_per_second": 68.779, | |
| "eval_steps_per_second": 2.612, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 24.857142857142858, | |
| "eval_accuracy": 0.7088607594936709, | |
| "eval_loss": 0.6648610234260559, | |
| "eval_runtime": 1.1222, | |
| "eval_samples_per_second": 70.395, | |
| "eval_steps_per_second": 2.673, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 25.714285714285715, | |
| "grad_norm": 2.387436866760254, | |
| "learning_rate": 3.888888888888889e-05, | |
| "loss": 0.2745, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "eval_accuracy": 0.7088607594936709, | |
| "eval_loss": 0.7311040163040161, | |
| "eval_runtime": 1.1344, | |
| "eval_samples_per_second": 69.64, | |
| "eval_steps_per_second": 2.645, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 26.857142857142858, | |
| "eval_accuracy": 0.7088607594936709, | |
| "eval_loss": 0.7192353010177612, | |
| "eval_runtime": 1.1573, | |
| "eval_samples_per_second": 68.263, | |
| "eval_steps_per_second": 2.592, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "eval_accuracy": 0.7215189873417721, | |
| "eval_loss": 0.7408412098884583, | |
| "eval_runtime": 1.1424, | |
| "eval_samples_per_second": 69.154, | |
| "eval_steps_per_second": 2.626, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 28.571428571428573, | |
| "grad_norm": 6.9804277420043945, | |
| "learning_rate": 3.7037037037037037e-05, | |
| "loss": 0.2494, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 28.857142857142858, | |
| "eval_accuracy": 0.810126582278481, | |
| "eval_loss": 0.5842322111129761, | |
| "eval_runtime": 1.1414, | |
| "eval_samples_per_second": 69.212, | |
| "eval_steps_per_second": 2.628, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "eval_accuracy": 0.759493670886076, | |
| "eval_loss": 0.5948711037635803, | |
| "eval_runtime": 1.1375, | |
| "eval_samples_per_second": 69.448, | |
| "eval_steps_per_second": 2.637, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 30.857142857142858, | |
| "eval_accuracy": 0.7468354430379747, | |
| "eval_loss": 0.6884945034980774, | |
| "eval_runtime": 1.1212, | |
| "eval_samples_per_second": 70.462, | |
| "eval_steps_per_second": 2.676, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 31.428571428571427, | |
| "grad_norm": 2.36922287940979, | |
| "learning_rate": 3.518518518518519e-05, | |
| "loss": 0.2291, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "eval_accuracy": 0.7088607594936709, | |
| "eval_loss": 0.874555766582489, | |
| "eval_runtime": 1.1468, | |
| "eval_samples_per_second": 68.887, | |
| "eval_steps_per_second": 2.616, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 32.857142857142854, | |
| "eval_accuracy": 0.7088607594936709, | |
| "eval_loss": 0.800529420375824, | |
| "eval_runtime": 1.1336, | |
| "eval_samples_per_second": 69.688, | |
| "eval_steps_per_second": 2.646, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "eval_accuracy": 0.7341772151898734, | |
| "eval_loss": 0.7033743858337402, | |
| "eval_runtime": 1.1296, | |
| "eval_samples_per_second": 69.939, | |
| "eval_steps_per_second": 2.656, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 34.285714285714285, | |
| "grad_norm": 3.042140483856201, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.2, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 34.857142857142854, | |
| "eval_accuracy": 0.7088607594936709, | |
| "eval_loss": 0.7047172784805298, | |
| "eval_runtime": 1.141, | |
| "eval_samples_per_second": 69.24, | |
| "eval_steps_per_second": 2.629, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "eval_accuracy": 0.7341772151898734, | |
| "eval_loss": 0.8362169861793518, | |
| "eval_runtime": 1.1313, | |
| "eval_samples_per_second": 69.828, | |
| "eval_steps_per_second": 2.652, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 36.857142857142854, | |
| "eval_accuracy": 0.7468354430379747, | |
| "eval_loss": 0.8509089946746826, | |
| "eval_runtime": 1.1603, | |
| "eval_samples_per_second": 68.085, | |
| "eval_steps_per_second": 2.586, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 37.142857142857146, | |
| "grad_norm": 2.2762393951416016, | |
| "learning_rate": 3.148148148148148e-05, | |
| "loss": 0.1674, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 38.0, | |
| "eval_accuracy": 0.759493670886076, | |
| "eval_loss": 0.9237213134765625, | |
| "eval_runtime": 1.1366, | |
| "eval_samples_per_second": 69.505, | |
| "eval_steps_per_second": 2.639, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 38.857142857142854, | |
| "eval_accuracy": 0.759493670886076, | |
| "eval_loss": 0.7526997327804565, | |
| "eval_runtime": 1.1354, | |
| "eval_samples_per_second": 69.579, | |
| "eval_steps_per_second": 2.642, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "grad_norm": 4.063281059265137, | |
| "learning_rate": 2.962962962962963e-05, | |
| "loss": 0.1764, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "eval_accuracy": 0.7468354430379747, | |
| "eval_loss": 0.7904393076896667, | |
| "eval_runtime": 1.1403, | |
| "eval_samples_per_second": 69.279, | |
| "eval_steps_per_second": 2.631, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 40.857142857142854, | |
| "eval_accuracy": 0.759493670886076, | |
| "eval_loss": 0.7332668304443359, | |
| "eval_runtime": 1.1312, | |
| "eval_samples_per_second": 69.835, | |
| "eval_steps_per_second": 2.652, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 42.0, | |
| "eval_accuracy": 0.7341772151898734, | |
| "eval_loss": 0.7777941226959229, | |
| "eval_runtime": 1.1452, | |
| "eval_samples_per_second": 68.983, | |
| "eval_steps_per_second": 2.62, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 42.857142857142854, | |
| "grad_norm": 1.630341649055481, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 0.1706, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 42.857142857142854, | |
| "eval_accuracy": 0.7721518987341772, | |
| "eval_loss": 0.7341672778129578, | |
| "eval_runtime": 1.1386, | |
| "eval_samples_per_second": 69.385, | |
| "eval_steps_per_second": 2.635, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 44.0, | |
| "eval_accuracy": 0.7468354430379747, | |
| "eval_loss": 0.8143826127052307, | |
| "eval_runtime": 1.1491, | |
| "eval_samples_per_second": 68.751, | |
| "eval_steps_per_second": 2.611, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 44.857142857142854, | |
| "eval_accuracy": 0.759493670886076, | |
| "eval_loss": 0.8299198150634766, | |
| "eval_runtime": 1.1355, | |
| "eval_samples_per_second": 69.576, | |
| "eval_steps_per_second": 2.642, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 45.714285714285715, | |
| "grad_norm": 3.097490072250366, | |
| "learning_rate": 2.5925925925925925e-05, | |
| "loss": 0.1617, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 46.0, | |
| "eval_accuracy": 0.7468354430379747, | |
| "eval_loss": 1.0110950469970703, | |
| "eval_runtime": 1.179, | |
| "eval_samples_per_second": 67.003, | |
| "eval_steps_per_second": 2.544, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 46.857142857142854, | |
| "eval_accuracy": 0.759493670886076, | |
| "eval_loss": 0.8602201342582703, | |
| "eval_runtime": 1.1263, | |
| "eval_samples_per_second": 70.14, | |
| "eval_steps_per_second": 2.664, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 48.0, | |
| "eval_accuracy": 0.7341772151898734, | |
| "eval_loss": 0.8332175016403198, | |
| "eval_runtime": 1.1297, | |
| "eval_samples_per_second": 69.928, | |
| "eval_steps_per_second": 2.656, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 48.57142857142857, | |
| "grad_norm": 4.115355014801025, | |
| "learning_rate": 2.4074074074074074e-05, | |
| "loss": 0.1622, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 48.857142857142854, | |
| "eval_accuracy": 0.7468354430379747, | |
| "eval_loss": 0.8297150731086731, | |
| "eval_runtime": 1.1338, | |
| "eval_samples_per_second": 69.677, | |
| "eval_steps_per_second": 2.646, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "eval_accuracy": 0.759493670886076, | |
| "eval_loss": 0.881697952747345, | |
| "eval_runtime": 1.1442, | |
| "eval_samples_per_second": 69.042, | |
| "eval_steps_per_second": 2.622, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 50.857142857142854, | |
| "eval_accuracy": 0.759493670886076, | |
| "eval_loss": 0.8742266297340393, | |
| "eval_runtime": 1.1356, | |
| "eval_samples_per_second": 69.566, | |
| "eval_steps_per_second": 2.642, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 51.42857142857143, | |
| "grad_norm": 2.3206353187561035, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 0.1437, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 52.0, | |
| "eval_accuracy": 0.759493670886076, | |
| "eval_loss": 1.0696383714675903, | |
| "eval_runtime": 1.1494, | |
| "eval_samples_per_second": 68.729, | |
| "eval_steps_per_second": 2.61, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 52.857142857142854, | |
| "eval_accuracy": 0.759493670886076, | |
| "eval_loss": 0.9411976337432861, | |
| "eval_runtime": 1.1462, | |
| "eval_samples_per_second": 68.926, | |
| "eval_steps_per_second": 2.617, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 54.0, | |
| "eval_accuracy": 0.7974683544303798, | |
| "eval_loss": 0.7411226630210876, | |
| "eval_runtime": 1.132, | |
| "eval_samples_per_second": 69.787, | |
| "eval_steps_per_second": 2.65, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 54.285714285714285, | |
| "grad_norm": 3.764564037322998, | |
| "learning_rate": 2.037037037037037e-05, | |
| "loss": 0.1492, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 54.857142857142854, | |
| "eval_accuracy": 0.759493670886076, | |
| "eval_loss": 0.9043496251106262, | |
| "eval_runtime": 1.1547, | |
| "eval_samples_per_second": 68.414, | |
| "eval_steps_per_second": 2.598, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 56.0, | |
| "eval_accuracy": 0.7848101265822784, | |
| "eval_loss": 0.7935737371444702, | |
| "eval_runtime": 1.1556, | |
| "eval_samples_per_second": 68.362, | |
| "eval_steps_per_second": 2.596, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 56.857142857142854, | |
| "eval_accuracy": 0.7721518987341772, | |
| "eval_loss": 0.8230910301208496, | |
| "eval_runtime": 1.1502, | |
| "eval_samples_per_second": 68.685, | |
| "eval_steps_per_second": 2.608, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 57.142857142857146, | |
| "grad_norm": 2.8445208072662354, | |
| "learning_rate": 1.8518518518518518e-05, | |
| "loss": 0.1279, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 58.0, | |
| "eval_accuracy": 0.7721518987341772, | |
| "eval_loss": 1.089436650276184, | |
| "eval_runtime": 1.168, | |
| "eval_samples_per_second": 67.635, | |
| "eval_steps_per_second": 2.568, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 58.857142857142854, | |
| "eval_accuracy": 0.7974683544303798, | |
| "eval_loss": 1.007054328918457, | |
| "eval_runtime": 1.1442, | |
| "eval_samples_per_second": 69.042, | |
| "eval_steps_per_second": 2.622, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 60.0, | |
| "grad_norm": 2.42126202583313, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.1317, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 60.0, | |
| "eval_accuracy": 0.7721518987341772, | |
| "eval_loss": 0.9893307089805603, | |
| "eval_runtime": 1.1414, | |
| "eval_samples_per_second": 69.214, | |
| "eval_steps_per_second": 2.628, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 60.857142857142854, | |
| "eval_accuracy": 0.7468354430379747, | |
| "eval_loss": 1.0475964546203613, | |
| "eval_runtime": 1.1369, | |
| "eval_samples_per_second": 69.487, | |
| "eval_steps_per_second": 2.639, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 62.0, | |
| "eval_accuracy": 0.7848101265822784, | |
| "eval_loss": 0.8080578446388245, | |
| "eval_runtime": 1.1418, | |
| "eval_samples_per_second": 69.191, | |
| "eval_steps_per_second": 2.628, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 62.857142857142854, | |
| "grad_norm": 2.587399482727051, | |
| "learning_rate": 1.4814814814814815e-05, | |
| "loss": 0.1456, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 62.857142857142854, | |
| "eval_accuracy": 0.7468354430379747, | |
| "eval_loss": 0.8136084079742432, | |
| "eval_runtime": 1.1341, | |
| "eval_samples_per_second": 69.657, | |
| "eval_steps_per_second": 2.645, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 64.0, | |
| "eval_accuracy": 0.7848101265822784, | |
| "eval_loss": 0.9612722992897034, | |
| "eval_runtime": 1.1822, | |
| "eval_samples_per_second": 66.825, | |
| "eval_steps_per_second": 2.538, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 64.85714285714286, | |
| "eval_accuracy": 0.7848101265822784, | |
| "eval_loss": 0.9783176183700562, | |
| "eval_runtime": 1.1425, | |
| "eval_samples_per_second": 69.145, | |
| "eval_steps_per_second": 2.626, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 65.71428571428571, | |
| "grad_norm": 2.009462594985962, | |
| "learning_rate": 1.2962962962962962e-05, | |
| "loss": 0.119, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 66.0, | |
| "eval_accuracy": 0.7721518987341772, | |
| "eval_loss": 1.0225824117660522, | |
| "eval_runtime": 1.1642, | |
| "eval_samples_per_second": 67.858, | |
| "eval_steps_per_second": 2.577, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 66.85714285714286, | |
| "eval_accuracy": 0.7721518987341772, | |
| "eval_loss": 1.0809653997421265, | |
| "eval_runtime": 1.1413, | |
| "eval_samples_per_second": 69.221, | |
| "eval_steps_per_second": 2.629, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 68.0, | |
| "eval_accuracy": 0.7974683544303798, | |
| "eval_loss": 0.9605960249900818, | |
| "eval_runtime": 1.1307, | |
| "eval_samples_per_second": 69.871, | |
| "eval_steps_per_second": 2.653, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 68.57142857142857, | |
| "grad_norm": 2.1945383548736572, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 0.1323, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 68.85714285714286, | |
| "eval_accuracy": 0.7848101265822784, | |
| "eval_loss": 0.9852112531661987, | |
| "eval_runtime": 1.1679, | |
| "eval_samples_per_second": 67.643, | |
| "eval_steps_per_second": 2.569, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 70.0, | |
| "eval_accuracy": 0.759493670886076, | |
| "eval_loss": 0.8826334476470947, | |
| "eval_runtime": 1.1435, | |
| "eval_samples_per_second": 69.086, | |
| "eval_steps_per_second": 2.624, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 70.85714285714286, | |
| "eval_accuracy": 0.7468354430379747, | |
| "eval_loss": 0.8169065117835999, | |
| "eval_runtime": 1.1429, | |
| "eval_samples_per_second": 69.122, | |
| "eval_steps_per_second": 2.625, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 71.42857142857143, | |
| "grad_norm": 4.010135650634766, | |
| "learning_rate": 9.259259259259259e-06, | |
| "loss": 0.126, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 72.0, | |
| "eval_accuracy": 0.759493670886076, | |
| "eval_loss": 0.8814924359321594, | |
| "eval_runtime": 1.1672, | |
| "eval_samples_per_second": 67.681, | |
| "eval_steps_per_second": 2.57, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 72.85714285714286, | |
| "eval_accuracy": 0.7721518987341772, | |
| "eval_loss": 0.9870919585227966, | |
| "eval_runtime": 1.1576, | |
| "eval_samples_per_second": 68.247, | |
| "eval_steps_per_second": 2.592, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 74.0, | |
| "eval_accuracy": 0.7721518987341772, | |
| "eval_loss": 0.8927366733551025, | |
| "eval_runtime": 1.1442, | |
| "eval_samples_per_second": 69.044, | |
| "eval_steps_per_second": 2.622, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 74.28571428571429, | |
| "grad_norm": 2.5069663524627686, | |
| "learning_rate": 7.4074074074074075e-06, | |
| "loss": 0.1013, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 74.85714285714286, | |
| "eval_accuracy": 0.7468354430379747, | |
| "eval_loss": 0.8364698886871338, | |
| "eval_runtime": 1.1466, | |
| "eval_samples_per_second": 68.9, | |
| "eval_steps_per_second": 2.616, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 76.0, | |
| "eval_accuracy": 0.7468354430379747, | |
| "eval_loss": 0.8423016667366028, | |
| "eval_runtime": 1.1449, | |
| "eval_samples_per_second": 69.0, | |
| "eval_steps_per_second": 2.62, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 76.85714285714286, | |
| "eval_accuracy": 0.7468354430379747, | |
| "eval_loss": 0.8330836296081543, | |
| "eval_runtime": 1.1475, | |
| "eval_samples_per_second": 68.845, | |
| "eval_steps_per_second": 2.614, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 77.14285714285714, | |
| "grad_norm": 2.3882863521575928, | |
| "learning_rate": 5.555555555555556e-06, | |
| "loss": 0.1142, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 78.0, | |
| "eval_accuracy": 0.7721518987341772, | |
| "eval_loss": 0.8203690052032471, | |
| "eval_runtime": 1.1426, | |
| "eval_samples_per_second": 69.14, | |
| "eval_steps_per_second": 2.626, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 78.85714285714286, | |
| "eval_accuracy": 0.7721518987341772, | |
| "eval_loss": 0.8286263942718506, | |
| "eval_runtime": 1.1509, | |
| "eval_samples_per_second": 68.642, | |
| "eval_steps_per_second": 2.607, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 80.0, | |
| "grad_norm": 2.164783239364624, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "loss": 0.1287, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 80.0, | |
| "eval_accuracy": 0.7721518987341772, | |
| "eval_loss": 0.8701770901679993, | |
| "eval_runtime": 1.1544, | |
| "eval_samples_per_second": 68.434, | |
| "eval_steps_per_second": 2.599, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 80.85714285714286, | |
| "eval_accuracy": 0.7721518987341772, | |
| "eval_loss": 0.9069644808769226, | |
| "eval_runtime": 1.1505, | |
| "eval_samples_per_second": 68.667, | |
| "eval_steps_per_second": 2.608, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 82.0, | |
| "eval_accuracy": 0.7721518987341772, | |
| "eval_loss": 0.9024941921234131, | |
| "eval_runtime": 1.1478, | |
| "eval_samples_per_second": 68.826, | |
| "eval_steps_per_second": 2.614, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 82.85714285714286, | |
| "grad_norm": 3.1293911933898926, | |
| "learning_rate": 1.8518518518518519e-06, | |
| "loss": 0.099, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 82.85714285714286, | |
| "eval_accuracy": 0.7721518987341772, | |
| "eval_loss": 0.8805702328681946, | |
| "eval_runtime": 1.1532, | |
| "eval_samples_per_second": 68.506, | |
| "eval_steps_per_second": 2.602, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 84.0, | |
| "eval_accuracy": 0.759493670886076, | |
| "eval_loss": 0.8637142777442932, | |
| "eval_runtime": 1.1416, | |
| "eval_samples_per_second": 69.202, | |
| "eval_steps_per_second": 2.628, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 84.85714285714286, | |
| "eval_accuracy": 0.759493670886076, | |
| "eval_loss": 0.8577674031257629, | |
| "eval_runtime": 1.1432, | |
| "eval_samples_per_second": 69.103, | |
| "eval_steps_per_second": 2.624, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 85.71428571428571, | |
| "grad_norm": 2.0774335861206055, | |
| "learning_rate": 0.0, | |
| "loss": 0.1141, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 85.71428571428571, | |
| "eval_accuracy": 0.7468354430379747, | |
| "eval_loss": 0.855110764503479, | |
| "eval_runtime": 1.1385, | |
| "eval_samples_per_second": 69.39, | |
| "eval_steps_per_second": 2.635, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 85.71428571428571, | |
| "step": 300, | |
| "total_flos": 2.9362240500074496e+18, | |
| "train_loss": 0.24900371452172598, | |
| "train_runtime": 1679.7368, | |
| "train_samples_per_second": 26.314, | |
| "train_steps_per_second": 0.179 | |
| }, | |
| { | |
| "epoch": 85.71428571428571, | |
| "eval_accuracy": 0.810126582278481, | |
| "eval_loss": 0.5842322111129761, | |
| "eval_runtime": 1.1419, | |
| "eval_samples_per_second": 69.185, | |
| "eval_steps_per_second": 2.627, | |
| "step": 300 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 300, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 100, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.9362240500074496e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |